The present invention relates generally to image compression. More particularly, the present invention relates image compression using neural networks.
Wavelet transforms are widely used in analysis, where they are known as xe2x80x9cmultiresolution analysisxe2x80x9d, and in image and audio compression, where they are used as a pyramid coding method for lossy compression. The wavelets used are generally from a very small set of analytically designed wavelets, such as Daubechies wavelets, or quadrature mirror filters (xe2x80x9cQMFxe2x80x9d). For some applications, designing specific wavelets with special coding properties would be beneficial.
Presently, compression methods directed to image and video compression attempt to minimize the amount of bandwidth used for a single band. There are no compression methods directed to reducing the amount of total activity in a network.
A method and apparatus of compressing data is described. The method and apparatus include constructing a neural network having a specific geometry using a finite and discrete Radon transform. The data is then fed through the neural network to produce a transformed data stream. The transformed data stream is thresholded. A fixed input signal is fed back through the neural network to generate a decoding calculation of an average value. The thresholded data stream is entropy encoded.